Skip to content

refactor: add TypeScript types and improve null handling in PDFParser… #403

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
modesty merged 1 commit into from
Sep 13, 2025
Merged

refactor: add TypeScript types and improve null handling in PDFParser… #403

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
76 changes: 47 additions & 29 deletions pdfparser.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,8 @@ import nodeUtil from "node:util";
import { readFile } from "node:fs/promises";
import { EventEmitter } from "node:events";
import { Buffer } from "node:buffer";
// eslint-disable-next-line @typescript-eslint/no-unused-vars
import { Readable } from "node:stream";

import PDFJS from "./lib/pdf.js";
import { ParserStream, StringifyStream } from "./lib/parserstream.js";
Expand Down Expand Up @@ -78,26 +80,34 @@ export default class PDFParser extends EventEmitter {
}

static #maxBinBufferCount = 10;
/** @type {Record<string, Buffer | null>} */
static #binBuffer = {};
static #instanceCounter = 0;

#password = "";
#context = null; // service context object, only used in Web Service project; null in command line #pdfFilePath = null;
#pdfFilePath = null; //current PDF file to load and parse, null means loading/parsing not started #data = null;
#pdfFileMTime = null; // last time the current pdf was modified, used to recognize changes and ignore cache #PDFJS = null;
#data = null; //if file read success, data is PDF content; if failed, data is "err" object #processFieldInfoXML = false;
/** @type {import('./src/types/pdfparser.js').PDFParserContext|null} */
#context = null; // service context object, only used in Web Service project; null in command line
/** @type {string|null} */
#pdfFilePath = null; //current PDF file to load and parse, null means loading/parsing not started
/** @type {number|null} */
#pdfFileMTime = null; // last time the current pdf was modified, used to recognize changes and ignore cache
/** @type {object|null} */
#data = null; //if file read success, data is PDF content; if failed, data is "err" object
/** @type {import('./lib/pdf.js').default|null} */
#PDFJS = null; //will be initialized in constructor
#processFieldInfoXML = false; //disable additional _fieldInfo.xml parsing and merging (do NOT set to true)

/**
* PDFParser constructor.
* @constructor PDFParser class.
* @param {object} context - The context object (only used in Web Service project); null in command line
* @param {import('./src/types/pdfparser.js').PDFParserContext|null} context - The context object (only used in Web Service project); null in command line
* @param {boolean} needRawText - Whether raw text is needed or not
* @param {string} password - The password for PDF file
* @info Private methods accessible using the [funcName].call(this, ...) syntax
*/
constructor(context, needRawText, password) {
super();
PDFParser.#instanceCounter++;
this.#context = context;
this.#pdfFilePath = null; //current PDF file to load and parse, null means loading/parsing not started this.#pdfFileMTime = null;
this.#pdfFileMTime = null; // last time the current pdf was modified, used to recognize changes and ignore cache this.#data = null;
Expand All @@ -109,20 +119,18 @@ export default class PDFParser extends EventEmitter {
}

/**
* @private
* @param {object} data - The parsed data
*/
#onPDFJSParseDataReady(data) {
if (!data) {
nodeUtil.p2jinfo("PDF parsing completed.");
this.emit("pdfParser_dataReady", this.#data);
} else {
this.#data = { ...this.#data, ...data };
this.#data = { ...(this.#data || {}), ...data };
}
}

/**
* @private
* @param {Error} err - The error object
*/
#onPDFJSParserDataError(err) {
Expand All @@ -131,11 +139,16 @@ export default class PDFParser extends EventEmitter {
}

/**
* @private
* @param {Buffer} buffer - The PDF buffer
* @param {Buffer|null} buffer - The PDF buffer
*/
#startParsingPDF(buffer) {
this.#data = {};
#startParsingPDF(buffer = null) {
this.#data = null;

if (!this.#PDFJS) {
this.#onPDFJSParserDataError(new Error("PDFJS parser not initialized"));
return;
}

this.#PDFJS.on("pdfjs_parseDataReady", (data) =>
this.#onPDFJSParseDataReady(data)
);
Expand All @@ -155,7 +168,6 @@ export default class PDFParser extends EventEmitter {
}

/**
* @private
* @returns {boolean}
*/
#processBinaryCache() {
Expand All @@ -166,7 +178,7 @@ export default class PDFParser extends EventEmitter {

const allKeys = Object.keys(PDFParser.#binBuffer);
if (allKeys.length > PDFParser.#maxBinBufferCount) {
const idx = this.id % PDFParser.#maxBinBufferCount;
const idx = PDFParser.#instanceCounter % PDFParser.#maxBinBufferCount;
const key = allKeys[idx];
PDFParser.#binBuffer[key] = null;
delete PDFParser.#binBuffer[key];
Expand All @@ -190,6 +202,9 @@ export default class PDFParser extends EventEmitter {
* @returns {string} The binBufferKey
*/
get binBufferKey() {
if (this.#pdfFilePath === null || this.#pdfFileMTime === null) {
return "";
}
return this.#pdfFilePath + this.#pdfFileMTime;
}

Expand All @@ -205,6 +220,7 @@ export default class PDFParser extends EventEmitter {
* Asynchronously load a PDF from a file path.
* @param {string} pdfFilePath - Path of the PDF file
* @param {number} verbosity - Verbosity level
* @returns {Promise<void>} Promise that resolves when PDF is loaded
*/
async loadPDF(pdfFilePath, verbosity) {
nodeUtil.verbosity(verbosity || 0);
Expand All @@ -214,7 +230,7 @@ export default class PDFParser extends EventEmitter {

try {
this.#pdfFileMTime = fs.statSync(pdfFilePath).mtimeMs;
if (this.#processFieldInfoXML) {
if (this.#processFieldInfoXML && this.#PDFJS) {
this.#PDFJS.tryLoadFieldInfoXML(pdfFilePath);
}

Expand Down Expand Up @@ -253,36 +269,36 @@ export default class PDFParser extends EventEmitter {
* @returns {string} Raw text content
*/
getRawTextContent() {
return this.#PDFJS.getRawTextContent();
return this.#PDFJS?.getRawTextContent() || "";
}

/**
* Retrieve raw text content stream.
* @returns {Stream} Raw text content stream
* @returns {Readable} Raw text content stream
*/
getRawTextContentStream() {
return ParserStream.createContentStream(this.getRawTextContent());
}

/**
* Retrieve all field types.
* @returns {object[]} All field types
* @returns {import('./src/types/pdfparser.js').FieldType[]} All field types
*/
getAllFieldsTypes() {
return this.#PDFJS.getAllFieldsTypes();
return this.#PDFJS?.getAllFieldsTypes() || [];
}

/**
* Retrieve all field types.
* @returns {object[]} All field types
* Retrieve all field data.
* @returns {import('./src/types/pdfparser.js').FieldType[]} All field data
*/
getAllFieldData() {
return this.#PDFJS.getAllFieldData();
return this.#PDFJS?.getAllFieldData() || [];
}

/**
* Retrieve all field types stream.
* @returns {Stream} All field types stream
* @returns {Readable} All field types stream
*/
getAllFieldsTypesStream() {
return ParserStream.createContentStream(this.getAllFieldsTypes());
Expand All @@ -293,12 +309,12 @@ export default class PDFParser extends EventEmitter {
* @returns {object} Merged text blocks
*/
getMergedTextBlocksIfNeeded() {
return this.#PDFJS.getMergedTextBlocksIfNeeded();
return this.#PDFJS?.getMergedTextBlocksIfNeeded() || {};
}

/**
* Retrieve merged text blocks stream.
* @returns {Stream} Merged text blocks stream
* @returns {Readable} Merged text blocks stream
*/
getMergedTextBlocksStream() {
return ParserStream.createContentStream(this.getMergedTextBlocksIfNeeded());
Expand All @@ -309,8 +325,9 @@ export default class PDFParser extends EventEmitter {
* @param {boolean} needRawText - Whether raw text is needed or not
*/
resetPDFJS(needRawText){
this.#PDFJS.destroy();
this.#PDFJS=new PDFJS(needRawText);
this.#PDFJS?.destroy();
this.#PDFJS = new PDFJS(needRawText);
PDFParser.#instanceCounter++;
}

/**
Expand All @@ -322,7 +339,7 @@ export default class PDFParser extends EventEmitter {

//context object will be set in Web Service project, but not in command line utility
if (this.#context) {
this.#context.destroy();
this.#context.destroy?.();
this.#context = null;
}

Expand All @@ -331,7 +348,8 @@ export default class PDFParser extends EventEmitter {
this.#data = null;
this.#processFieldInfoXML = false; //disable additional _fieldInfo.xml parsing and merging (do NOT set to true)

this.#PDFJS.destroy();
this.#PDFJS?.destroy();
this.#PDFJS = null;
PDFParser.#instanceCounter--;
}
}
6 changes: 5 additions & 1 deletion src/types/pdfparser.d.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@ export declare class PDFParser extends EventEmitter {
// eslint-disable-next-line @typescript-eslint/naming-convention
static get _PARSER_SIG(): string;

constructor(context?: any, needRawText?: boolean, password?: string);
constructor(context?: PDFParserContext | null, needRawText?: boolean, password?: string);
on<K extends keyof EventMap>(eventName: K, listener: EventMap[K]): this;

readonly data: object | null;
Expand Down Expand Up @@ -172,4 +172,8 @@ export declare interface Box {
style?: number;
}

export interface PDFParserContext {
destroy?(): void;
}

export default PDFParser
Loading